Clinical research
Association of RANTES G-403A gene polymorphism
with increased risk of coronary arteriosclerosis
Eleonora Simeoni
a,b,1, Bernhard R. Winkelmann
c,1, Michael M. Hoffmann
d,1,
Sylvain Fleury
a,b, Juan Ruiz
e, Lukas Kappenberger
a, Winfried M€
arz
f,
Giuseppe Vassalli
a,g,*
aDivision of Cardiology, University of Lausanne Medical School, CHUV-BH-10, 1011 Lausanne, Switzerland b
Experimental Surgery, University of Lausanne Medical School, Lausanne, Switzerland c
The Cooperation Unit Pharmacogenomics/Applied Genomics, University of Heidelberg, Heidelberg, Germany dThe Department of Clinical Chemistry, University of Freiburg, Freiburg, Germany
e
Endocrinology, University of Lausanne Medical School, Lausanne, Switzerland fThe Department of Clinical Chemistry, Graz, Austria
g
The Institute of Microbiology, University of Lausanne Medical School, Lausanne, Switzerland
Received 27 November 2003; revised 30 April 2004; accepted 5 May 2004
See page 1378 for the editorial comment on this article
Aims Polymorphisms in the RANTES (G-403A), monocyte chemoattractant protein-1 (MCP-1; A-2518G), stromal cell-derived factor-1b (SDF-1b; G801A), and C–C chemo-kine receptor-5 (CCR5; D32) genes have been associated with functional effects. These chemokines have been implicated in leucocyte recruitment to arterial lesions. In a case-control study, we explored relations between these polymorphisms and coronary artery disease (CAD), with respect to angiographic abnormalities and acute coronary syndromes (ACS).
Methods and Results The LUdwigshafen Risk and Cardiovascular health (LURIC) co-hort was genotyped by RFLP-PCR. Based on coronary angiography, individuals were sub-divided into CAD casesðn ¼ 2694Þ and controls ðn ¼ 530Þ. RANTES-403 genotype frequencies were significantly different in cases and controlsðv2¼ 4:17; p ¼ 0:041Þ, as were A allele carrier frequencies (36.01% vs. 30.19%, OR¼ 1.30 [95%-CI¼ 1.06–1.60], p ¼ 0:010). By multivariate analysis, RANTES A-403 retained signifi-cant association with CADðv2¼ 8:40; p ¼ 0:0038Þ. RANTES A-403 was associated with increased ACS prevalence (OR¼ 1.36 [95%-CI ¼ 1.08–1.71], p ¼ 0:0073). MCP-1 G-2518, SDF-1b A801, and CCR5 D32 were not associated with CAD.
Conclusions RANTES A-403 was associated with CAD independently from conventional risk factors and CRP or fibrinogen as inflammatory biomarkers. The association was enhanced in smokers and ACS, conditions where platelet activation and inflammation predominate. RANTES A-403 may increase genetic susceptibility to CAD.
c 2004 The European Society of Cardiology. Published by Elsevier Ltd. All rightsreserved. KEYWORDS RANTES; MCP-1; SDF-1b; CCR5; Polymorphism; Coronary artery disease
*Corresponding author. Tel.:þ41-21-3140076; fax: þ41-21-3140013. E-mail address:gvassall@hospvd.ch(G. Vassalli).
1These authors contributed equally to this work.
doi:10.1016/j.ehj.2004.06.035.
0195-668X/$ - see front matter
c 2004 The European Society of Cardiology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ehj.2004.05.005Introduction
Arteriosclerosis is a multi-factorial condition which is determined both by environmental and genetic factors, and involves a strong inflammatory component.1 Cell adhesion and signalling interactions between circulating leucocytes and the endothelium are key events in the recruitment of leucocytes to inflammatory foci. While beneficial in defense against infection and cancer, in-teractions between leucocytes and endothelium initiate arterial lesion formation. In response to chemoattractant signals, monocytes and T-cells adhere to inflamed en-dothelium, transmigrate into the sub-endothelial space, recognise intra-lesional antigens such as oxidised low-density lipoproteins (ox-LDL), and acquire an activated phenotype in the vessel wall.2
Chemoattractant peptides (chemokines) regulate leucocyte trafficking in inflammatory diseases.3The C–C chemokine Regulated upon Activation, Normal T-cell Expressed and Secreted (RANTES), monocyte chemo-attractant protein-1 (MCP-1), and stromal cell-derived factor-1b (SDF-1bÞ chemokines have been implicated in atherogenesis. RANTES is a potent chemoattractant for monocytes, lymphocytes, eosinophils, and basophils. RANTES expression has been demonstrated in T-cells in human atherosclerotic plaques,4 as well as in lympho-cytes, macrophages, endothelial cells, and vascular smooth muscle cells in transplant-associated arterio-sclerosis.5 Mice deficient in the RANTES gene show im-paired T-cell and monocyte recruitment to inflammatory sites.6MCP-1 is a potent chemoattractant for monocytes and memory T-cells, and an activator for monocytes/ macrophages within arterial lesions.7 Mice deficient in either the MCP-1 or in the C–C chemokine receptor (CCR)-2 gene that encode the main receptor for MCP-1 show reduced arterial lesion formation when crossed with arteriosclerosis-prone mice deficient in the LDL receptor gene.8;9These findings support a central role for MCP-1 in atherogenesis. SDF-1b is a potent platelet ag-onist that has been implicated in lymphocyte arrest on inflamed endothelium.10
Recently, common polymorphisms in the RANTES, MCP-1, SDF-1b, and CCR5 genes have been associated with functional and biological effects. The A-403 variant in the promoter region of the RANTES gene resulted in up to 8-fold increased constitutive transcriptional ac-tivity after transient transfection of the human mast-cell line HMC-1 and the T-mast-cell line Jurkat with reporter vectors driven by either the mutant or the reference RANTES promoter.11 Clinical associations of RANTES A-403 with inflammatory diseases such as atopic derma-titis,11;12 asthma,12 polymyalgia rheumatica,13 sarcoido-sis,14diabetic nephropathy,15as well as susceptibility to HIV infection and disease progression16 have been reported.
The G-2518 variant in the distal regulatory region of the MCP-1 gene resulted in increased MCP-1 expression in human peripheral blood mononuclear cells (PBMC) by 2-fold upon IL-1 stimulation.17An association between MCP-1 G-2518 homozygosity and coronary artery disease (CAD) was also reported recently.18
The A801 variant in the 30 untranslated region of the SDF-1b gene has been associated with genetic restriction of AIDS pathogenesis,19as well as with enhanced CD34þ progenitor cell mobilisation from the bone marrow.20 Although it has been proposed that the A801 variant may enhance SDF-1b bioavailability,19 available data on its effect on SDF-1b expression are conflicting.21–23
The D32 deletion mutation in the gene encoding CCR5, a receptor for RANTES, results in a frameshift and premature termination of transcript translation, pre-venting expression of the receptor molecule on cell surfaces.24 Associations of CCR5 D32 with resistance to HIV infection,24reduced severity of asthma25 and rheu-matoid arthritis,26as well as decreased risk of premature myocardial infarction have been reported.27
These findings suggest functional gene variants in chemokines and chemokine receptors may affect ge-netic susceptibility to inflammatory diseases, most likely by modulating leucocyte recruitment and acti-vation in inflammatory foci. We hypothesised that functional polymorphisms in the RANTES, MCP-1, SDF-1b, and CCR5 genes influence the development of coronary artery lesions. To test this hypothesis, we genotyped a large cohort studied by coronary angiog-raphy, and we correlated genotypic and angiographic findings. A significant association with angiographically detectable coronary artery lesions was found for RANTES G-403A, but not for the other candidate gene polymorphisms.
Methods
Study population
We genotyped the LUdwigshafen Risk and Cardiovascular health (LURIC) cohort of 3316 individuals who underwent coronary angiography because of chest pain or non-invasive tests consistent with myocardial ischaemia.28All individuals were Caucasians born in Germany from par-ents of German ancestry. Three thousand two hundred and twenty four subjects were included in the analysis after exclusion of 92 individuals due to incomplete an-giographic data ðn ¼ 49Þ, clinical evidence of arterial disease including stroke, carotid artery stenosis and se-vere peripheral artery disease (defined by a history of intermittent claudication, angiographic documentation of lumen obstruction, aneurysm of the abdominal aorta, or a history of peripheral arterial intervention for ath-erosclerotic disease) in patients with normal coronary angiogramsðn ¼ 39Þ, a history of myocardial infarction in patients with normal coronary angiograms ðn ¼ 2Þ, or missing genetic dataðn ¼ 2Þ.
Angiographic classification
Angiographic criteria defining CAD cases vs. controls were as follows: (1) lumen reduction P 20% on one or more major epicardial coronary arteries; (2) lumen re-duction 10–19% in three or more (out of 15) coronary artery segments. Mild coronary lesions were defined as
<50% lumen narrowing and moderate to severe coronary artery stenoses as P 50% lumen narrowing.
Clinical data
Clinical characteristics of cases and controls are shown in Table 1. Diabetes was defined according to the new ADA and WHO criteria29;30by presence of one or more of the following criteria: (1) treatment with hypoglycaemic agents; (2) two diagnostic tests showing high blood glu-cose levels (fasting plasma gluglu-cose P 126 mg/dL and/or 2-h plasma glucose P 200 mg/dL during an oral glucose tolerance test according to the WHO criteria30); (3) hi-story of diabetes confirmed by at least one diagnostic test performed during the study. Subjectsðn ¼ 97Þ with a single high blood glucose level (no confirmatory test) were not assigned an unequivocal diabetes status. Hy-pertension was defined as a supine systolic blood pres-sure P 140 mm Hg and/or diastolic prespres-sure P 90 mm Hg (average from three consecutive measurements after 10 min of rest). Hypercholesterolaemia was defined as fasting cholesterol P 240 mg/dL or treatment with lipid-lowering drugs. High triglycerides (TG) were P 150 mg/mL. Low HDL-cholesterol (HDL-C) was <35 mg/mL in males and \45 mg/mL in females. Fibrinogen was measured by the STA fibrinogen/STA Stago Kit (Stago Diagnostica/Roche). C-reactive protein (CRP) was mea-sured using a mAb nephelometric assay (N LATEX CRP mono; Dade Behring). Among the 3224 individuals in-cluded in the analysis, data sampling was incomplete with respect to TG and HDL-C in 1 subject each, and to fibrinogen, CRP, and waist-to-hip ratio (WHR) in 3, 6, and 40 patients, respectively.
Acute coronary syndromes
In addition to an analysis of the presence of any angio-graphic lesion (CAD) as the primary endpoint, the preva-lence of acute coronary syndromes (ACS) was analysed as a secondary endpoint reflecting functionally active le-sions resulting in major coronary events in CAD patients. ACS were defined as unstable angina or acute myocardial infarction, with or without ST-segment elevation, in
ac-cordance with the Joint European Society of Cardiology/ American College of Cardiology Committee.31
Genotyping
Genotyping of RANTES G-403A, MCP-1 A-2518G, and SDF-1b G801A polymorphisms was performed by restriction fragment length polymorphism (RFLP)-PCR as de-scribed,13;17;19 with a minor modification for RANTES G-403A. This modification consisted of the change of a single base (capital letter, underlined) within the 3’ primer (50-gttcctgcttattcattacagatcGta-30) to create a new restriction digestion site for better discrimination of the PCR fragments. Because genotype frequencies of RANTES-403 did not comply with Hardy–Weinberg equi-librium (HWE) proportions in the control group (see: “Results”), genotyping was repeated twice in this group using higher concentrations of restriction enzyme to rule out partial digestion, and using an improved method for RFLP with a modified 50 primer (50 -caatgcccagc-tcagatcaactgcctc-30). Genotyping of CCR5 D32 was performed by PCR, as described.24 Technicians and in-vestigators were blinded with respect to case-control status. Two investigators scored gels independently and unclear positions were repeated. Genotype results were typed twice into data files to avoid errors in recording.
Statistical analysis
Power calculations with a pre-set level of significance ða ¼ 0:05Þ had been calculated for equal numbers of cases and controls according to Fleiss et al.,32 in the LURIC project (see LURIC study design,28 Appendix II). For an unequal number of casesðn ¼ 2694Þ and controls ðn ¼ 530Þ, the power to detect a 6% or 7% difference in genotype/allele frequencies for a two-sided test with a level of significance of p¼ 0:05 was 74% or 86%, re-spectively. Data are shown as frequencies or means±standard deviation (SD). CRP did not have a normal standard distribution and was log-transformed for analysis. The JMP program (version 5; SAS Institute, Cary, NC) was used for statistical analysis. Frequencies were compared with the Pearson v2-test; means were
com-Table 1 Clinical characteristics of controls and cases, and their associations with CAD (univariate analysis)
Characteristic n Controls, n¼ 530 Cases, n¼ 2694 p-value
Age: years 3224 56.9±11.9 63.8±9.9 \0.0001 Male gender, (%) 3224 51.3 73.9 \0.0001 Smoker, (%) 3224 46.4 67.7 \0.0001 Arterial hypertension, (%) 3224 42.1 61.9 \0.0001 Type 2 diabetes, (%) 3127 7.5 22.5 \0.0001 Hypercholesterolaemia, (%) 3224 36.6 68.4 \0.0001 Hypertriglyceridaemia, (%) 3223 39.1 48.8 \0.0001 Low HDL cholesterol, (%) 3223 44.3 55.2 \0.0001 Waist-to-hip ratio 3184 0.93±0.08 0.97±0.07 \0.0001 Fibrinogen: mg/dL 3221 347.5±74.8 406.4±109.9 \0.0001 C reactive protein: mg/L 3218 335 (295–392) 388 (327–463) \0.0001
Nominal variables (definitions: see “Methods”) are percentages. Age, WHR, and fibrinogen are mean values±SD; CRP are median values (25–75 percentiles). Missing data in controls for diabetesðn ¼ 8Þ, WHR ðn ¼ 5Þ, and fibrinogen ðn ¼ 1Þ. Missing data in cases for diabetes ðn ¼ 89Þ, TG ðn ¼ 1Þ, HDL-C ðn ¼ 1Þ, WHR ðn ¼ 35Þ, fibrinogen ðn ¼ 2Þ, and CRP ðn ¼ 6Þ.
pared with the two-tailed unpaired Student’s t-test or ANOVA test. Gene counting method estimated allele frequencies, and the v2 analysis was used to test for deviations of genotype proportions from HWE. The Ar-mitage’s test for trend in proportions (does not assume HWE) was used for statistical analysis of genotype and allele carrier frequencies according to Sasieni,33as pro-posed by Xu et al.34In addition, the variance of differ-ences in allele frequencies was determined using the method by Schaid and Jacobsen.35 Odds ratios for ge-notype analysis were calculated according to Sasieni.33 Correction for multiple testing was performed by Bon-ferroni adjustment. The Cochran–Mantel–Haenszel method was used to test association of RANTES A-403 with CAD after stratification for cigarette smoking, age, and gender. The independent contribution of RANTES A-403 to CAD was determined by nominal logistic analysis, with CAD as the dependent variable, using three differ-ent models with increasingly conservative character: (1) conventional coronary risk factors including age, male gender, smoking, hypertension, diabetes, hypercho-lesterolaemia, low HDL-C, high TG, and WHR were en-tered as co-variates; (2) in addition to those traditional risk factors, CRP was entered as a co-variate; (3) in ad-dition to traad-ditional risk factors and CRP, fibrinogen was also entered as a co-variate. Age, WHR, CRP, and fi-brinogen were analysed as continuous variables that satisfied the assumption of linearity. This assumption was checked by: (1) sub-dividing each continuous variable into five sub-groups, corresponding to the each 20th percentiles of the distribution; (2) calculating the logits of CAD probabilities for each sub-group; (3) analysing the linearity of the logit vs. the sub-divided variable by simple linear regression (age, r2¼ 0:95; WHR, r2¼ 0:91; CRP, r2¼ 0:99; fibrinogen, r2¼ 0:97). To test the influ-ence of gene–gene interaction of MCP-1 A-2518G, SDF-1b G801A, and CCR5 D32 polymorphisms with RANTES G-403A and CAD, a multivariate model with interaction terms (MCP-1 A-2518G*RANTES G-403A, SDF-1b G801A*RANTES G-403A, and CCR5 D32*RANTES G-403A)
was used. We adopted a significance level of p < 0:05 for this exploratory study.
Results
Angiographic results
Based on coronary angiographic criteria, 3224 subjects of the LURIC cohort included in the analysis were sub-divided into casesðn ¼ 2694Þ with abnormal angiograms and controlsðn ¼ 530Þ with normal angiograms. Among cases, the majorityðn ¼ 2231Þ showed moderate to se-vere lumen reduction ( P 50%) in one ðn ¼ 620Þ, two ðn ¼ 623Þ, or three ðn ¼ 988Þ major epicardial coronary arteries. A minority of cases presented with 0-vessel disease (<50% stenosis; n¼ 463) defined as mild angio-graphic abnormalities with lumen reduction of 20–49% of at least one major coronary artery ðn ¼ 336Þ or minor narrowings (10–19%) of P 3 coronary artery segments ðn ¼ 127Þ.
Associations between clinical characteristics and CAD
Patients with CAD presented with the typical changes in the cardiovascular risk profile when compared to sub-jects without CAD (all p < 0:0001; Table 1).
Genotypic data and HWE
Genotype frequencies for RANTES-403 complied with HWE both in the study population as a whole and in cases (data not shown), but deviated from it in controls ðv2¼ 4:49; p ¼ 0:034Þ. The true Type-I error rate ðdÞ and the fraction of maximum discrepancy with respect to the HWE model (f, where HWE: f¼ 0; maximum discrep-ancy: f¼ 1:0), calculated according to Schaid and Jac-obsen,35 were modest ðd ¼ 0:00239; f ¼ 0:015Þ. Odds ratios for genotype analysis were calculated according to
Table 2 Clinical characteristics of controls and cases in carriers (AGþ AA) and non-carriers (GG) of RANTES A-403 Characteristics Controls,ðnÞ Cases,ðnÞ
GG (n¼ 370) AGþ AA (n ¼ 160) GG (n¼ 1724) AGþ AA (n ¼ 970) Age: years,ðnÞ 57.1±11.6 (370) 56.5±12.6 (160) 63.5±10.1 (1724) 64.3±9.5 (970) Male gender, %ðnÞ 51.1 (189) 51.9 (83) 75.1 (1294) 71.9 (697) Smoker, %ðnÞ 48.8 (179) 41.9 (67) 68.8 (1187) 65.8 (638) Arterial hypertension, %ðnÞ 42.7 (158) 40.6 (65) 61.9 (1068) 61.7 (599) Type 2 diabetes, %ðnÞ 6.0 (22) 10.8 (17) 21.8 (364) 23.7 (222) Hypercholesterolemia, %ðnÞ 37.8 (140) 33.7 (54) 68.7 (1184) 68.0 (660) Hypertriglyceridemia, %ðnÞ 39.5 (146) 38.1 (61) 50.0 (861) 46.6 (452) Low HDL cholesterol, %ðnÞ 43.2 (160) 46.9 (75) 55.5 (957) 54.6 (530) Waist to hip ratioðnÞ 0.93±0.08 (367) 0.93±0.08 (158) 0.97±0.07 (1707) 0.96±0.08 (952) Fibrinogen, mg/dLðnÞ 344.7±73.2 (369) 353.7±78.1 (160) 403.5±108.6 (1722) 411.5±112.0 (970) CRP, mg/L [25–75 percentiles]ðnÞ 335 [294–390] (370) 338 [297–396] (160) 384 [326–460] (1720) 393 [332–469] (968)
Age, WHR, and fibrinogen are mean values±SD; CRP are median values [25–75 percentiles]. Missing data in carriers of the A allele for diabetes ðn ¼ 2Þ and WHR ðn ¼ 2Þ in controls, and for diabetes ðn ¼ 33Þ, WHR ðn ¼ 18Þ, and CRP ðn ¼ 2Þ in cases. Missing data in non-carriers of the A allele for diabetesðn ¼ 6Þ, WHR ðn ¼ 3Þ, and fibrinogen ðn ¼ 1Þ in controls, and for diabetes ðn ¼ 56Þ, TG ðn ¼ 1Þ, low HDL-C ðn ¼ 1Þ, WHR ðn ¼ 17Þ, fibrinogen ðn ¼ 2Þ, and CRP ðn ¼ 4Þ in cases.
Sasieni.33Genotype frequencies for MCP-1-2518, SDF-1b 801, and CCR5 D32 were in HWE in both cases and con-trols (data not shown).
Association between RANTES A-403 and coronary risk factors
Clinical characteristics of carriers (AG/AA) and non-car-riers (GG) of the RANTES A-403 allele were comparable (p > 0:05; Table 2). In the study population as a whole, only plasma fibrinogen was correlated with RANTES-403 genotype and allele carrier frequencies (F ratio¼ 3.47, p¼ 0:0310 and t ¼ 2:59; p ¼ 0:0098, respectively).
Association between RANTES A-403 and CAD
Genotype and allele frequencies of RANTES-403, as well as allele carrier frequencies of RANTES A-403, were sig-nificantly different between cases and controls (Table 3). The analysis of allelic frequencies using the method by Schaid and Jacobsen35 that corrects for deviations from HWE yielded Z¼ 2:04 ðp ¼ 0:041Þ. For comparison, assuming a HWE model would yield ZHWE¼ 2:06 ðp ¼ 0:040Þ, consistent with a minor difference between the HWE and non-HWE models for this analysis. After sub-division of cases based on coronary lesion severity, allele carrier frequencies of RANTES A-403 were significantly increased both in cases with moderate to severe lesions ( P 50%) and in those with mild lesions (<50%), as com-pared to controls (Table 4). Data stratification for gender and age revealed no significant effect of these parame-ters on the association between RANTES A-403 and CAD (data not shown). After data stratification for cigarette smoking, the significant association between RANTES A-403 and CAD was slightly enhanced (genotype and A allele carrier frequencies: v2¼ 9:41; p ¼ 0:0090 and v2¼ 8:20; p¼ 0:0042, respectively).
RANTES A-403 is an independent predictor of CAD
Genotype and allele frequencies of RANTES A-403 re-tained significant association with CAD, independent of conventional risk factors, in a multivariate model with conventional coronary risk factors entered as co-variates (first model; Table 5). The association was maintained when CRP was entered as a co-variate in addition to conventional risk factors (second model; v2¼ 6:65; p¼ 0:036 and v2¼ 6:47; p ¼ 0:011 for genotype and allele frequencies, respectively), or when CRP and fi-brinogen were both entered as additional co-variates (third model; v2¼ 6:19; p ¼ 0:045 and v2¼ 6:02; p ¼ 0:014 for genotype and allele frequencies, respectively).
Association between RANTES A-403 and ACS
Among CAD cases, ACS occurred in 1032 patients, whereas 1662 patients did not suffer from ACS. Genotype frequencies of RANTES-403 were significantly different
Ta ble 3 Genot ype, allele, and allele carrier fre quencies o f RANTE S A-403 in cont rols and cases G roup Genotypes frequencies , % ðn Þ A allele frequencies , % A allele carrier fre quen cies, % GG AG AA v 2 p OR Zp O R (95% CI ) v 2 p OR (95% CI) Cont rols 69.8 (370) 25.8 (137) 4.3 (23) 4.17 0.041 1.14 17.3 2.04 0.041 1.20 (1.01–1.4 3) 30.2 6.58 0.010 1.30 (1.06–1.6 0) Cases 64.0 (1724) 32.0 (862) 4.0 (108) 20.0 36.0 (Calculation o f OR, see 33 ).
between cases with ACS and controls ðv2¼ 5:17; OR¼ 1:19; p ¼ 0:023Þ, but not between cases without ACS and controlsðv2¼ 2:69; OR ¼ 1:10; p ¼ 0:10Þ. Simi-larly, allele carrier frequencies of RANTES A-403 were significantly increased in cases with ACS compared with controls (37.02% vs. 30.19%; v2¼ 7:20; OR ¼ 1:36 [95%-CI¼ 1.08–1.71], p ¼ 0:0073), even after Bonferroni ad-justment for multiple testing. In contrast, A allele carrier frequencies were only slightly increased in cases without ACS (35.38%; v2¼ 4:82; OR ¼ 1:27 [95%-CI ¼ 1.02–1.57], p¼ 0:028).
Lack of association between MCP-1 G-2518, SDF-1b A801, CCR5 D32, and CAD
Genotype, allele, and allele carrier frequencies of MCP-1-2518, SDF-1b 801, and CCR5-32 were not significantly different in cases and controls (Table 6). In disagreement with a previous report,27 CCR5 D32 was not correlated with premature myocardial infarction (data not shown). No significant gene–gene interaction of the four poly-morphisms with respect to CAD risk was observed (data not shown).
Table 4 Genotype and allele frequencies of RANTES A-403 with respect to CAD severity
Group Genotypes frequencies, %ðnÞ A allele carrier frequencies, %
GG AG AA v2 p OR v2 p OR (95% CI) Controls 69.8 (370) 25.8 (137) 4.3 (23) 30.2 CAD\50% 61.6 (285) 33.7 (156) 4.7 (22) 5.67 0.017 1.24 38.4 7.50 0.006 1.44 (1.10–1.90) CAD P 50% 64.5 (1439) 31.6 (706) 3.9 (86) 3.16 0.075 1.11 35.5 5.35 0.021 1.27 (1.03–1.57)
Data are shown for subsets with mild (lumen reduction: <50%) or moderate to severe CAD (lumen reduction: P 50%)(Calculation of OR, see33).
Table 6 Genotype, allele, and allele carrier frequencies of MCP-1 A-2518G, SDF-1b G801A, and CCR5 D32 gene polymorphisms in controls and cases
Genotype frequencies, n (%) Allele 2* frequencies, %
Allele 2* carrier frequencies, % Polymorphism Group Genotype 1–1* Genotype 1–2* Genotype 2–2*
MCP-1 A-2518G Controls 288 (54.5) 207 (39.2) 33 (6.2) 25.8 45.4 Cases 1423 (52.8) 1068 (39.7) 202 (7.5) 27.3 47.2 SDF-1b G808A Controls 350 (66.2) 159 (30.1) 20 (3.8) 18.8 33.8 Cases 1709 (63.5) 885 (32.9) 96 (3.6) 20.0 36.5 CCR5 D32 Controls 423 (80.1) 99 (18.7) 6 (1.1) 10.5 19.9 Cases 2093 (78.1) 553 (20.6) 35 (1.3) 11.6 21.9
(*) 1, frequent allele and 2, rare allele. Missing data for MCP-1 A-2518G: n¼ 2 controls and 1 case; SDF-1b G808A: n ¼ 1 control and four cases; CCR5 D32: n¼ 2 controls and 13 cases. p > 0:10 in cases vs. controls for each polymorphism with respect to differences in genotype, allele, and allele carrier frequencies between cases and controls.
Table 5 Nominal logistic fit, with CAD as the dependent variable (first model; see “Methods”) Genotype analysis A allele carrier analysis
Source of variation DF R2 L-R v2 p value DF R2 L-R v2 p value
Multivariate model (n¼ 3086) All variables 11 0.24 665.9 <0.0001 10 0.24 665.6 <0.0001 Age (yr) 1 156.1 <0.0001 1 156.2 <0.0001 Gender 1 60.8 <0.0001 1 60.8 <0.0001 Smoker 1 61.7 <0.0001 1 61.7 <0.0001 Arterial hypertension 1 23.6 <0.0001 1 23.6 <0.0001 Type 2 diabetes 1 23.6 <0.0001 1 23.6 <0.0001
Waist to hip ratio 1 1.98 0.16 1 2.03 0.15
Hypercholesterolemia 1 153.1 <0.0001 1 153.2 <0.0001 Hypertriglyceridemia 1 0.43 0.51 1 0.44 0.51 Low HDL cholesterol 1 21.2 <0.0001 1 21.2 <0.0001 RANTES-403 polymorphism 2ð#Þ 8.64 0.013 1ðÞ 8.40 0.0038
Type of analysis:ð#Þ3-class (GG¼ 0, AG ¼ 1, and AA ¼ 2);ðÞ2-class (GG¼ 0 and AG + AA ¼ 1). Due to incomplete data in a small number of subjects
Discussion
Gene polymorphisms that modify expression and/or bio-availability of chemokines and their cellular receptors may affect leucocyte trafficking in inflammatory dis-eases, including arteriosclerosis. In our case-control study of n¼ 3224 subjects, RANTES A-403, but not the other three gene variants studied, was significantly as-sociated with CAD. Carriers of the A allele were at in-creased risk of developing angiographically detectable coronary lesions compared with individuals homozygous for the G allele. The association was stronger in patients with mild (<50%) coronary lesions ðp ¼ 0:017Þ and in those with ACSðp ¼ 0:023Þ. After Bonferroni adjustment for multiple testing of the four gene polymorphisms, the frequency of carriers of the A allele remained signifi-cantly increased in CAD patients. The association be-tween RANTES A-403 and CAD was enhanced in cigarette smokers. Finally, RANTES A-403 remained significantly associated with CAD after multivariate adjustment for conventional risk factors and for CRP and fibrinogen, two biomarkers of systemic inflammation associated with CAD.36;37
Thus, we were able to replicate recent results from Hungary showing an increased frequency of the RANTES A-403 allele in patients undergoing coronary artery by-pass surgery.18 However, the statistical significance in this smaller cohort ðn ¼ 628Þ was borderline (p¼ 0:052; OR ¼ 1:33 [95%-CI ¼ 1.00–1.77]). Of note, this odds ratio is closely comparable to the one obtained in the present study. Obviously, replicating results in independent cohorts is crucial for establishing a genetic association.
The biological mechanism underlying this genetic as-sociation most likely is increased gene expression from the RANTES A-403 promoter variant.11 RANTES expres-sion is differentially regulated in many tissues. Several putative cis-acting elements have been described in the RANTES promoter region.38The A-403 variant introduces a new consensus binding site for the GATA transcription factor family. This variant has been associated with en-hanced RANTES promoter activity in human T-cell and mast cell lines, which express GATA binding proteins, but not in epithelial cells, which do not express these pro-teins.11 Because megakaryocytes constitutively express both RANTES and GATA binding proteins,39 RANTES pro-duction and release by platelets might be enhanced in carriers of the A-403 allele. RANTES stored in platelet secretory vesicles is released upon platelet activation and immobilised on the surface of inflamed endothe-lium,40–42 where it promotes shear-resistant monocyte arrest and trans-endothelial migration,40;43–45 while generating a critical signal for chemokine production by monocytes.46RANTES deposition on the luminal surface of carotid arteries has been demonstrated in apolipo-protein E-deficient mice with early atherosclerotic le-sions in vivo, as well as in human atherectomy samples.40 The RANTES antagonist Met-RANTES inhibited monocyte recruitment on carotid endothelium, neointima forma-tion and macrophage accumulaforma-tion after arterial injury, as well as atherosclerosis progression in uninjured
ar-teries in apolipoprotein E-deficient mice.40;46;47 These findings suggest an important role for RANTES in vascular disease, potentially explaining the enhanced association between RANTES A-403 and CAD in conditions charac-terised by platelet activation, such as smoking and ACS. Recently, RANTES A-403 has been associated with several inflammatory diseases.11–15 Polymorphisms in other chemokine genes have also been associated with CAD. Fraktalkine and its receptor, CX3CR1, which are expressed in human plaques,48 have been implicated in atherogenesis.49An association between the I249 variant of CX3CR1 and reduced endothelial dysfunction and CAD lesions was reported.50 The loss-of-function mutation CX3CR1-M280 was associated with cardiovascular pro-tection in Framingham participants.51 The rare Val64Ile polymorphism in the CCR2 gene coding for the main MCP-1 receptor was associated with reduced coronary calci-fications in one study,52and with premature myocardial infarction (but not coronary lesions) in another study.53 As mentioned, MCP-1 G-2518 homozygosity was also as-sociated with CAD.18 However, this association was not replicated in the present study in a larger cohort, nor was the reported protection against premature myocardial infarction by CCR5 D32.27
A methodological aspect of the present study that deserves discussion, like in many case-control associa-tion studies, is the appropriateness of the control group. Because angiography is relatively insensitive for early lesions that do not impact on the vascular lumen, early lesions undetectable by angiography could not be ruled out in controls, nor was coronary microvascular dys-function. Nevertheless, coronary angiography remains a gold standard for the diagnosis of CAD, and it is well known that the future incidence of cardiac events is low in subjects with normal coronary angiograms. Another issue that deserves consideration is the departure of RANTES-403 genotypes from HWE proportions in controls ðv2¼ 4:49; p ¼ 0:034Þ, the number of AA homozygous subjects being higher than expected. Apparent violations of HWE might arise because of errors in genotyping. A literature survey found that 16 out of 133 single nucle-otide polymorphisms (12%) from 75 case-control studies deviated from HWE.34 Therefore, genotyping was re-peated three times (with two different methods) to carefully rule out genotyping errors in controls. Another potential cause for deviations from HWE is ethnic di-versity within the population. The LURIC study was re-stricted to native Germans of German ancestry living in the Rhine valley area (a region where >95% are Cauca-sians of German descent and where second and third-generation family members are often alive) in order to limit ethnic heterogeneity and to enhance access to family members. All other gene polymorphisms tested in this cohort in the present and previous studies were in HWE, also consistent with limited ethnic heterogeneity. Thus, no explanation could be found for the departure from HWE in controls. A comparison of our data with those from the Hungarian cohort18(complying with HWE) shows almost identical frequencies of control subjects homozygous for the frequent G allele (69.1% vs. 69.8% in LURIC) and, hence, of controls carrying the A allele.
Stated differently, frequencies of carriers of the A allele in controls, which significantly differed from those in cases in our study ðv2¼ 6:58; p ¼ 0:010Þ, were closely comparable in the two studies. It also should be em-phasised that this analysis, like the one of genotype frequencies, was performed using statistical tests that do not assume HWE.33Moreover, allelic frequencies were analysed using the test by Schaid and Jacobsen35 that corrects for deviations from HWE. This test also revealed that the true Type-I error rate and the fraction of max-imum discrepancy with the HWE model were modest, corroborating the validity of our statistical data.
In conclusion, the RANTES A-403 allele was associ-ated with CAD independently from conventional car-diovascular risk factors and from CRP and fibrinogen, two biomarkers of vascular inflammation. The associa-tion was enhanced in smokers and ACS, two condiassocia-tions that are accompanied by platelet activation and vascu-lar inflammation. Our findings are consistent with bio-logical effects of Met-RANTES in experimental models of arteriosclerosis.40;46;47 Novel non-peptide, orally bio-available RANTES antagonists that have been success-fully tested in heart transplantation54and HIV models55 might also prove effective in preventing cardiovascular disease.
Acknowledgements
This work was funded by the Teo Rossi di Montelera Foundation, Lausanne, Switzerland, the Swiss Cardiology Foundation and the Emma Muschamp Foundation, Lau-sanne, Switzerland.
References
1. Hansson GK. Immune mechanisms in atherosclerosis. Arterioscler Thromb Vasc Biol 2001;21:1876–90.
2. Stemme S, Faber B, Holm J et al. T lymphocytes from human atherosclerotic plaques recognize oxidized low-density lipoprotein. Proc Natl Acad Sci USA 1995;92:3893–7.
3. Baggiolini M. Chemokines and leukocyte traffic. Nature 1998;392:565–8.
4. Wilcox JN, Nelken NA, Coughlin SR et al. Local expression of inflammatory cytokines in human atherosclerotic plaques. J Arte-rioscler Thromb 1994;1(Suppl):S3–S10.
5. Pattison JM, Nelson PJ, Huie P et al. RANTES chemokine expression in transplant-associated accelerated atherosclerosis. J Heart Lung Transplant 1996;15:1194–9.
6. Makino Y, Cook DN, Smithies O et al. Impaired T cell function in RANTES-deficient mice. Clin Immunol 2002;102:302–9.
7. Yla-Herttuala S, Lipton BA, Rosenfeld ME et al. Expression of monocyte chemoattractant protein 1 in macrophage-rich areas of human and rabbit atherosclerotic lesions. Proc Natl Acad Sci USA 1991;88:5252–6.
8. Boring L, Gosling J, Cleary M et al. Decreased lesion formation in CCR2= mice reveals a role for chemokines in the initiation of atherosclerosis. Nature 1998;394:894–7.
9. Gu L, Okada Y, Clinton SK et al. Absence of monocyte chemoattr-actant protein-1 reduces atherosclerosis in low density lipoprotein receptor-deficient mice. Mol Cell 1998;2:275–81.
10. Abi-Younes S, Sauty A, Mach F et al. The stromal cell-derived factor-1 chemokine is a potent platelet agonist highly expressed in athero-sclerotic plaques. Circ Res 2000;86:131–8.
11. Nickel RG, Casolaro V, Wahn U et al. Atopic dermatitis is associated with a functional mutation in the promoter of the C–C chemokine RANTES. J Immunol 2000;164:1612–6.
12. Fryer AA, Spiteri MA, Bianco A et al. The )403 G ! A promoter
polymorphism in the RANTES gene is associated with atopy and asthma. Genes Immun 2000;1:509–14.
13. Makki RF, al Sharif F, Gonzalez-Gay MA et al. RANTES gene polymorphism in polymyalgia rheumatica, giant cell arteritis and rheumatoid arthritis. Clin Exp Rheumatol 2000;18:391–3. 14. Takada T, Suzuki E, Ishida T et al. Polymorphism in RANTES
chemokine promoter affects extent of sarcoidosis in a Japanese population. Tissue Antigens 2001;58:293–8.
15. Nakajima K, Tanaka Y, Nomiyama T et al. RANTES promoter genotype is associated with diabetic nephropathy in type 2 diabetic subjects. Diabetes Care 2003;26:892–8.
16. McDermott DH, Beecroft MJ, Kleeberger CA et al. Chemokine RANTES promoter polymorphism affects risk of both HIV infection and disease progression in the Multicenter AIDS Cohort Study. AIDS 2000;14:2671–8.
17. Rovin BH, Lu L, Saxena R. A novel polymorphism in the MCP-1 gene regulatory region that influences MCP-1 expression. Biochem Biophys Res Commun 1999;259:344–8.
18. Szalai C, Duba J, Prohaszka Z et al. Involvement of polymorphisms in the chemokine system in the susceptibility for coronary artery disease (CAD). Coincidence of elevated Lp (a) and MCP-1-2518 G/G genotype in CAD patients. Atherosclerosis 2001;158:233–9. 19. Winkler C, Modi W, Smith MW et al. Genetic restriction of AIDS
pathogenesis by an SDF-1 chemokine gene variant. Science 1998;279:389–93.
20. Benboubker L, Watier H, Carion A et al. Association between the SDF1-3’A allele and high levels of CD34(+) progenitor cells mobilized into peripheral blood in humans. Br J Haematol 2001;113:247–50. 21. Sei S, O’Neill DP, Stewart SK et al. Increased level of stromal
cell-derived factor-1 mRNA in peripheral blood mononuclear cells from children with AIDS-related lymphoma. Cancer Res 2001;61:5028–37. 22. Soriano A, Martinez C, Garcia F et al. Plasma stromal cell-derived factor (SDF)-1 levels, SDF1-3’A genotype, and expression of CXCR4 on T lymphocytes: their impact on resistance to human immunodefi-ciency virus type 1 infection and its progression. J Infect Dis 2002;186:922–31.
23. Arya SK, Ginsberg CC, Davis-Warren A et al. In vitro phenotype of SDF1 gene mutant that delays the onset of human immunodeficiency virus disease in vivo. J Hum Virol 1999;2:133–8.
24. Liu R, Paxton WA, Choe S et al. Homozygous defect in HIV-1 coreceptor accounts for resistance of some multiply-exposed indi-viduals to HIV-1 infection. Cell 1996;86:367–77.
25. Hall IP, Wheatley A, Christie G et al. Association of CCR5 delta-32 with reduced risk of asthma. Lancet 1999;354:1264–5.
26. Zapico I, Coto E, Rodriguez A et al. CCR5 (chemokine receptor-5) DNA-polymorphism influences the severity of rheumatoid arthritis. Genes Immun 2000;1:288–9.
27. Gonzalez P, Alvarez R, Batalla A et al. Genetic variation at the chemokine receptors CCR5/CCR2 in myocardial infarction. Genes Immun 2001;2:191–5.
28. Winkelmann BR, M€arz W, Boehm BO et al. Rationale and design of the LURIC study – a resource for functional genomics, pharmacogenom-ics and long-term prognosis of cardiovascular disease. Pharmacoge-nomics 2001;2(Suppl 1):S9–S73.
29. Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of the expert committee on the diagnosis and classification of diabetes mellitus [from the American Diabetes Association, ADA]. Diabetes Care 2002;25(Suppl 1):S1–20. 30. World Health Organization. Definition, diagnosis and classification of
diabetes mellitus and its complications. Report of a WHO consulta-tion. Part 1. Diagnosis and classification of diabetes mellitus WHO/ NCD/NCS/99. Geneva: World Health Organization; 1999.
31. The Joint European Society of Cardiology/American College of Cardiology Committee: myocardial infarction redefined – a consen-sus document of the Joint European Society of Cardiology/American College of Cardiology Committee for the redefinition of myocardial infarction. Eur Heart J 2000;21:1502–13.
32. Fleiss JL, Tytun A, Ury HK. A simple approximation for calculating sample sizes for comparing independent proportions. Biometrics 1980;36:343–6.
33. Sasieni PD. From genotypes to genes: doubling the sample size. Biometrics 1997;53:1253–61.
34. Xu J, Turner A, Little J et al. Positive results in association studies with departure from Hardy–Weinberg equilibrium: hint for genotyp-ing error. Hum Genet 2002;111:573–4.
35. Schaid DJ, Jacobsen SJ. Biased tests of association: comparisons of allele frequencies when departing from Hardy–Weinberg propor-tions. Am J Epidemiol 1999;149:706–11.
36. Ridker PM, Cushman M, Stampfer MG et al. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med 1997;336:973–9.
37. Kannel WB, Wolf PA, Castelli WP et al. Fibrinogen and risk of cardiovascular disease. The Framingham Study. JAMA 1987;258:1183–6.
38. Nelson PJ, Kim HT, Manning WC et al. Genomic organization and transcriptional regulation of the RANTES chemokine gene. J Immunol 1993;151:2601–12.
39. Orkin SH. GATA-binding transcription factors in hematopoietic cells. Blood 1992;80:575–81.
40. von Hundelshausen P, Weber KSC, Huo Y et al. RANTES deposition by platelets triggers monocyte arrest on inflamed and atherosclerotic endothelium. Circulation 2001;103:1772–7.
41. Lindemann S, Tolley ND, Dixon DA et al. Activated platelets mediate inflammatory signaling by regulated interleukin 1 beta synthesis. J Cell Biol 2001;154:485–90.
42. Weyrich AS, Prescott SM, Zimmerman GA. Platelets, endothelial cells, inflammatory chemokines, and restenosis. Circulation 2002;106:1433–5.
43. Weber C, Weber KS, Klier C et al. Specialized roles of the chemokine receptors CCR1 and CCR5 in the recruitment of monocytes and T (H)1-like/CD45RO(+) T cells. Blood 2001;97:1144–6.
44. Wiedermann CJ, Kowald E, Reinisch N et al. Monocyte hapoptaxis induced by RANTES chemokine. Curr Biol 1993;3:735– 9.
45. Schober A, Manka D, von Hundelshausen P et al. Deposition of platelet RANTES triggering monocyte recruitment requires P-selectin
and is involved in neointima formation after arterial injury. Circu-lation 2002;106:1523–9.
46. Weyrich AS, Elstad MR, McEver RP et al. Activated platelets signal chemokine synthesis by human monocytes. J Clin Invest 1996;97:1525–34.
47. Veillard NR, Kwak B, Pelli G et al. Antagonism of RANTES receptors reduces atherosclerotic plaque formation in mice. Circ Res 2004;94:253–61.
48. Wong BW, Wong D, McManus BM. Characterization of fractalkine (CX3CL1) and CX3CR1 in human coronary arteries with native atherosclerosis, diabetes mellitus, and transplant vascular disease. Cardiovasc Pathol 2002;11:332–8.
49. Lesnik P, Haskell CA, Charo IF. Decreased atherosclerosis in CX3CR1= mice reveals a role for fractalkine in atherogenesis. J Clin Invest 2003;111:333–40.
50. McDermott DH, Halcox JPJ, Schenke WH et al. Association between polymorphism in the chemokine receptor CX3CR1 and coronary vascular endothelial dysfunction and atherosclerosis. Circ Res 2001;89:401–7.
51. McDermott DH, Fong AM, Yang Q et al. Chemokine receptor mutant CX3CR1-M280 has impaired adhesive function and correlates with protection from cardiovascular disease in humans. J Clin Invest 2003;111:1241–50.
52. Valdes AM, Wolfe ML, O’Brien EJ et al. Val64Ile polymorphism in the C–C chemokine receptor 2 is associated with reduced coronary artery calcification. Arterioscler Thromb Vasc Biol 2002;22:1924–8. 53. Ortlepp JR, Vesper K, Mevissen V et al. Chemokine receptor (CCR2) genotype is associated with myocardial infarction and heart failure in patients under 65 years of age. J Mol Med 2003;81:363–7. 54. Horuk R, Clayberger C, Krensky AM et al. A non-peptide functional
antagonist of the CCR1 chemokine receptor is effective in rat heart transplant rejection. J Biol Chem 2001;276:4199–204.
55. Strizki JM, Xu S, Wagner NE et al. SCH-C (SCH 351125), an orally bioavailable, small molecule antagonist of the chemokine receptor CCR5, is a potent inhibitor of HIV-1 infection in vitro and in vivo. Proc Natl Acad Sci USA 2001;98:12718–23.
